The differentiation of mesenchymal stem cells to vascular cells regulated by the HMGB1/RAGE axis: its application in cell therapy for transplant arteriosclerosis
Mesenchymal stem cell (MSC) transplantation shows promise for treating transplant arteriosclerosis, at least partly via promoting endothelial regeneration. However, the efficacy and safety are still under investigation especially regarding recent findings that neointimal smooth muscle cells are derived from MSC-like cells. The high mobility group box 1 (HMGB1)/receptor for advanced glycation end-product (RAGE) axis is involved in regulating proliferation, migration, and differentiation of MSCs, and therefore it can be presumably applied to improve the outcome of cell therapy. The aim of the current study was to investigate this hypothesis.
Rat MSCs were treated with HMGB1 or modified with HMGB1 vectors to activate the HMGB1/RAGE axis. RAGE was targeted and inhibited by specific short hairpin RNA vectors. We assessed the capacity for cell proliferation, migration, and differentiation after vector transfection in vitro and in a rat model of transplant arteriosclerosis. The expression of CD31 and α-smooth muscle actin (αSMA) was determined to evaluate the differentiation of MSCs to endothelial cells and smooth muscle cells.
Exogenous HMGB1 treatment and transfection with HMGB1 vectors promoted MSC migration and vascular endothelial growth factor (VEGF)-induced differentiation to CD31+ cells while inhibiting their proliferation and platelet-derived growth factor (PDGF)-induced differentiation to αSMA+ cells. Such an effect was blocked by RAGE knockdown. HMGB1-modified cells preferably migrated to graft neointima and differentiated to CD31+ cells along with significant relief of transplant arteriosclerosis and inhibition of HMGB1 and RAGE expression in graft vessels. RAGE knockdown inhibited cell migration to graft vessels.
HMGB1 stimulated MSCs to migrate and differentiate to endothelial cells via RAGE signaling, which we translated to successful application in cell therapy for transplant arteriosclerosis.
Dulbecco’s modified Eagle’s medium
Fetal bovine serum
Green fluorescent protein
High mobility group box 1
Mesenchymal stem cell
Nuclear factor κB
Platelet-derived growth factor
Receptor for advanced glycation end-product
Cherry red fluorescent protein
Short hairpin RNA
Smooth muscle antigen
Smooth muscle cell
Vascular endothelial growth factor
Despite the development of surgical techniques and new immune suppressive agents, chronic allograft rejection remains an obstacle to long-term allograft survival . Transplant arteriosclerosis (TA) as a specific form of arteriosclerosis is typically evident in chronically rejected organs. The affected arteries show a thickening of the intimal layers that are filled with vascular smooth muscle cells (SMCs) and extracellular matrix. The process named as intimal hyperplasia or neointimal formation gives rise to arterial stenosis which restricts the blood supply to grafts with a consequent late graft loss. Therefore, it makes sense to explore effective interventions for TA.
Transplantation of mesenchymal stem cells (MSCs) was introduced to prolong allograft survival with satisfactory outcomes in preclinical and clinical studies [2, 3, 4, 5, 6, 7]. The therapeutic effects were initially linked to the immunomodulatory properties of MSCs, including induction of regulatory T cells, secretion of anti-inflammatory cytokines, and suppression of alloantigen reactive lymphocytes. Further research was undertaken to generate durable chimerism and induce immune tolerance by MSC-based therapy [8, 9], although this turned out to be difficult. Other studies revealed that MSC transplantation was effective in treating arteriosclerosis. Neointimal formation was attenuated by MSC transplantation in balloon-induced arterial injury models, which was associated with enhanced endothelial repair [10, 11]. Moreover, transplantation of endothelial-like cells derived from MSCs preferably suppressed intimal hyperplasia following vascular injury . This suggested that MSCs attenuated arteriosclerosis at least partly via endothelial regeneration. However, the safety of MSC-based therapy was queried in recent studies on the origin of neointimal SMCs. Traditionally, it was believed that the key process of neointimal formation included the proliferation and migration of medial SMCs which switched from the contractile to the proliferative or synthetic phenotype in response to vascular injury. But it has now been revealed that multipotent stem cells which reside in vascular walls migrate to the intimal layers of injured vessels and subsequently differentiate into neointimal SMCs [13, 14]. Although the stem cells exist physiologically as a small population, they are capable of self-renewal and proliferation, and some of them acquire an MSC-like phenotype. Whether resident vascular stem cells contribute more than medial SMCs to intimal hyperplasia is still in dispute, however, and further investigation is still ongoing. Nevertheless, this discovery raised concerns about the participation of transplanted MSCs in arteriosclerosis presumably via SMC differentiation.
In this scenario, we have been exploring ways of improving the efficacy of MSC transplantation with a special focus on the regulation of cell differentiation that would promote endothelial regeneration. The current study aimed to evaluate whether the differentiation of MSCs to vascular cells would be influenced by high mobility group box 1 (HMGB1) and whether it could be used to provide favorable outcomes of MSC-based therapy for transplant arteriosclerosis. HMGB1 is an evolutionarily abundant and highly conserved DNA binding protein, and is passively released from necrotic cells or actively secreted by inflammatory cells to the extracellular milieu upon cell stress to regulate innate and adaptive immunity. HMGB1 stimulates the immune response, at least in part, through interaction with its principal binding partner, receptor for advanced glycation end-product (RAGE). Both HMGB1 and RAGE are highly expressed in atherosclerotic plaques to accelerate the progression of atherosclerosis . Moreover, the HMGB1/RAGE axis is instrumental in promoting acute and chronic rejection [16, 17, 18, 19, 20, 21, 22, 23]. Despite its proinflammatory properties, HMGB1 promotes tissue repair by recruiting stem cells. For instance, HMGB1 promoted the proliferation, migration, and osteogenic differentiation of MSCs via STAT3 and Ras/MAPK signaling pathways [24, 25]. Cell migration was also enhanced by HMGB1-induced synthesis of the CXCL12 receptor CXCR4, which possessed potent chemotactic activity for MSCs . However, whether HMGB1 regulates the differentiation of MSCs to vascular cells was not investigated, and data on the proliferation and migration of MSCs induced by HMGB1 were not consistent. Two recent studies revealed that extracellular HMGB1 released from necrotic cells and activated platelets inhibited MSC migration to apoptotic cardiac myocytes mediated by hepatocyte growth factor and its receptor MET [27, 28]. Another study demonstrated that RAGE knockout rescued streptozotocin-induced suppression of MSC proliferation and differentiation . Overall, previous studies seem to obtain mutually exclusive results at least partly due to different experimental conditions. Nevertheless, they suggest a diversity and complexity of MSC response to HMGB1 stimulation. Therefore, it is necessary to investigate the role of the HMGB1/RAGE axis on MSC transplantation for TA.
Cell culture and reagents
F344 rat bone marrow MSCs were purchased from Cyagen Biosciences Company (Shanghai, China) and cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS; Gibco, ThermoFisher Scientific). The cells were grown as an adherent culture at 37 °C in a humidified atmosphere of 5% CO2 and 95% air. Vascular endothelial growth factor (VEGF) and platelet-derived growth factor (PDGF)-BB were obtained from ProSpec (St. Louis, MO, USA). RIPA lysis buffer, BCA protein assay kit, and CCK-8 kit were purchased from Beyotime Biotech (Shanghai, China). Antibodies together with their manufacturers are listed below: rabbit anti-HMGB1 IgG, rabbit anti-RAGE IgG, mouse anti-CD31 IgG, mouse anti-α-smooth muscle antigen (αSMA) IgG, mouse anti-CD68 IgG, phycoerythrin (PE)-conjugated and fluorescein isothiocyanate (FITC)-conjugated anti-αSMA antibodies, secondary antibodies of Alexa Fluor 488 conjugated goat anti-mouse IgG, Alexa Fluor 488 conjugated goat anti-rabbit IgG, Alexa Fluor 647 conjugated goat anti-mouse IgG and Alexa Fluor 647 conjugated goat anti-rabbit IgG, horseradish peroxidase (HRP) conjugated goat anti-rabbit IgG and HRP-conjugated goat anti-mouse IgG (all from Abcam, Shanghai, China), PE-conjugated and FITC-conjugated anti-CD31 antibodies (Miltenyi Biosciences, Shanghai, China), mouse anti-rat tubulin IgG (BD, Shanghai, China), and mouse anti-FLAG IgG (Cell Signaling Technology, Shanghai, China). Recombinant HMGB1 protein and other reagents were purchased from Sigma-Aldrich unless otherwise stated.
All lentiviral vectors were constructed by Genechem (Shanghai, China): HMGB1 expression vector pLV-HMGB1 and its negative control pLV-control, which contained the reporter gene green fluorescent protein (GFP), a set of three RAGE short hairpin RNA (shRNA) vectors including shRNA-RAGE1, shRNA-RAGE2, and shRNA-RAGE3, plus one negative control pshRNA-control which contained the sequence of cherry red fluorescent protein (RFP) (Additional file 1: Supplemental Materials and Methods). Before viral transfection, MSCs were plated in six-well plates at 2 × 105 cells per well and cultured for 24 h to allow cell adhesion. The virus was then added at a concentration of 6 × 105 transduction units per well to the culture medium supplemented with 5 μg/ml polybrene to enhance transfection efficiency. After transfection for 12 h, the medium was replaced with fresh medium, and the cells were cultured for 3 days. The transfected cells were selected by 5 μg/ml puromycin and later propagated in culture medium. Transfection efficiency was assessed by calculating the percentage of fluorescent protein-labeled cells in the total population (Additional file 2: Figure S1). QRT-PCR revealed that the transfected cells retained MSC phenotypes of CD29 and CD90, which was similar to untransfected MSCs (Additional file 1: Supplemental Materials and Methods; Additional file 3: Figure S2).
Animal model, study groups, and stem cell transplantation
Animal groups and treatment protocols
Infusion of pLV-control-transfected cells
Infusion of pLV-HMGB1-transfected cells
Infusion of pshRNA-control-transfected cells
Infusion of pshRNA-RAGE1-transfected cells
Infusion of pshRNA-RAGE3-transfected cells
Cell proliferation was measured by a CCK-8 kit. Firstly, the cells were seeded in 96-well plates at 4 × 103 cells per well. After incubation for 24 h, the cells were cultured in the medium containing 10% CCK-8 solution for 2 h. Finally, the optical density (OD) was detected at 450 nm. The experiments were performed in triplicate for each study group to calculate the average OD value. The cell survival in the treatment groups was represented by the OD value normalized to that of untreated MSCs.
Transwell migration assay
Cell migration was measured by transwell assay. Firstly, the chambers of 24-well plates were separated into upper and lower compartments by transwell inserts with 8-μm pore membrane filters (Corning, Shanghai, China). Next, the cells were seeded in the upper compartment at 8 × 104 cells per well and incubated for 24 h. The upper compartment contained serum-free medium, while the medium supplemented with 15% FBS was added to the lower compartment to stimulate cell migration. Finally, the cells migrating to the lower side of the membrane were fixed in methanol, stained with 0.1% crystal violet solution, and visualized by phase-contrast microscopy. The experiments were repeated five times for each group, and data are expressed as the average migrating cell count per ×200 field micrograph.
In vitro differentiation of MSCs to vascular cells
The MSCs were cultured in DMEM supplemented with 5% FBS and 12.5 ng/ml PDGF-BB for 14 days to induce SMC differentiation. PDGF-BB was replaced by 25 ng/ml VEGF to induce endothelial differentiation. The expression of CD31 and αSMA was detected by flow cytometry to identify the cells that differentiated to endothelial cells (ECs) and SMCs, respectively.
Histological and immunohistochemical analysis
The cross sections of graft vessels were stained with hematoxylin and eosin to evaluate neointima formation. The neointimal thickness was normalized to the full thickness of vessel walls and expressed as a percentage. Other sections were stained by an immunoperoxidase technique to measure the expression of HMGB1 and RAGE. Briefly, the sections were processed through antigen retrieval and blocking procedures and then consecutively incubated with primary antibody, HRP-conjugated secondary antibody, and the substrate solution containing 3,3-diaminobenzidine. The nuclei were counterstained with Mayer’s hemalum solution.
Evans blue staining
The aortic endothelial barrier integrity was assessed by the Evans blue dye exclusion test as previously described . Briefly, freshly procured graft vessels were perfused with heparinized phosphate-buffered solution to remove blood cells, subsequently with 0.3% Evans blue dye for 5 min, and finally with phosphate-buffered solution to remove the dye residue. The vessels were opened by longitudinally cutting in half and then fixed between two glass slides. The images were photographed by Nikon Coolpix 4500. Positive staining indicated endothelium dysfunction or denudation. The area was quantified with image pro plus analysis software to calculate the ratio of unstained area (stained without dark blue) to the total intraluminal surface area of the graft vessels.
Cell and tissue immunofluorescence
The expression of HMGB1, RAGE, CD31, and αSMA in the cultured cells and graft vessels was detected by immunofluorescence. Firstly, frozen sections of graft vessels and cultured cells were fixed with cold acetone and blocked by 10% goat serum. Target protein was probed with specific primary antibodies and then visualized by incubation with corresponding fluorescent protein-labeled secondary antibodies. Finally, the nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI).
Western blot analysis
Total cellular protein was isolated from fresh tissue or cells with a RIPA lysis buffer containing protease inhibitor. Protein concentrations in the supernatants were measured by BCA protein assay. Protein samples were separated using 12% SDS-PAGE gels, transferred to polyvinylidene difluoride membranes, and blocked with 5% nonfat milk solution. The membranes were incubated with primary antibodies and corresponding HRP-conjugated secondary antibodies. Antibody binding was visualized using the ECL system.
CD31 and αSMA on the cell surface were detected by flow cytometry. Briefly, the cells were transferred to 5-ml flow tubes and incubated with PE- or FITC-conjugated antibodies in the dark for 30 min at 4 °C. The cells were then washed and suspended in phosphate-buffered saline. Fluorescence was quantified using BD FACSCanto II flow cytometer (BD Biosciences, CA, USA). The gates were set based on isotype control antibodies.
Statistical analysis of the data was performed using GraphPad Prism 5.0. Data are expressed as mean ± standard deviation and compared between groups by the Mann-Whitney test. A P value < 0.05 indicates a significant difference.
HMGB1 inhibited MSC proliferation, but promoted their migration through RAGE activation
HMGB1 promoted VEGF-induced differentiation of MSCs to CD31+ cells via RAGE activation
HMGB1 inhibited PDGF-induced differentiation of MSCs to αSMA+ cells via RAGE activation
Transplantation of pLV-HMGB1-transfected MSCs remarkably ameliorated TA along with inhibition of HMGB1 and RAGE expression in allograft vessels
pLV-HMGB1 transfection promoted MSC migration and endothelial differentiation in vivo and improved endothelial barrier integrity
TA is a common manifestation of chronically rejected grafts which appears with diffuse narrowing and occlusion of the graft vessels caused by concentric neointimal formation. Although it is regarded as a particular type of arteriosclerosis with significant differences to native atherosclerosis, many similarities exist regarding histological changes as well as signaling pathways that are involved, including the HMGB1/RAGE axis [20, 21, 23, 31, 32]. HMGB1 was upregulated in arteriosclerotic plaques and allografts to attract immune cells and stimulate innate and adaptive immunity [15, 23]. Ligation of RAGE by HMGB1 stimulated more production of HMGB1 and further activation of the HMGB1/RAGE axis, forming a vicious cycle to amplify the immune response . In the present study, both HMGB1 and RAGE were highly expressed in allograft vessels, especially in hyperplastic neointima where cytoplasmic translocation of HMGB1 was discovered, suggesting its extracellular release and signaling via RAGE. Despite its proinflammatory nature, some studies have revealed beneficial applications of HMGB1 in injury states. One such application was overexpression of HMGB1 or administration of exogenous HMGB1 to promote tissue remodeling and repair. Extracellular HMGB1 is a potent chemoattractant to recruit regenerative cells to damaged tissue. For instance, more endothelial progenitor cells were recruited and engrafted into ischemic myocardium with transgenic overexpression of HMGB1 than the mock control . The reactive astrocytes in the peri-infarct cortex upregulated HMGB1 which stimulated the migration of endogenous endothelial progenitor cells to promote neurovascular remodeling during stroke recovery . The chemotaxis was at least in part mediated by its principal receptor, RAGE [34, 35]. The present study revealed that MSC migration was enhanced by both exogenous HMGB1 treatment and pLV-HMGB1 transfection. However, the impact of HMGB1 on MSC proliferation is still controversial. An early study showed that HMGB1 inhibited MSC proliferation  while the opposite effect was evoked by stimulating MSCs with HMGB1 released from necrotic tumor cells . The present study found that cell growth was inhibited by exogenous HMGB1 treatment as well as pLV-HMGB1 transfection. The previous disagreement was presumably attributable to the variation in experimental conditions. Nevertheless, we demonstrated for the first time that HMGB1 displayed different interactions with VEGF and PDGF, eliciting opposite effects on differentiation of MSCs to ECs and SMCs. HMGB1 inhibited PDGF-induced differentiation of MSCs to SMCs, but it stimulated endothelial differentiation induced by VEGF. Intriguingly, HMGB1 treatment alone failed to induce phenotypic changes (Additional file 4: Figure S3), and there was barely a trace of VEGF or PDGF-BB in the culture supernatant. Similarly, HMGB1 promoted osteogenic differentiation of MSCs, but lost its function in the medium that was devoid of osteoinductive supplements . Thus, HMGB1 regulated MSC differentiation by interacting with other growth factors, and more specifically in our study it promoted VEGF-induced endothelial differentiation but inhibited the differentiation to the SMC phenotype induced by PDGF.
The present study further demonstrates that the regulation of HMGB1 on proliferation, migration, and differentiation of MSCs is through a RAGE-dependent pathway. Firstly, although HMGB1 impeded the growth of MSCs in a dose-dependent manner, the effect was greatly abolished by shRNA-mediated RAGE knockdown. Secondly, RAGE knockdown impaired the stimulatory effect of HMGB1 on endothelial differentiation of MSCs and abolished its inhibition of differentiation to SMCs. Thirdly, RAGE knockdown inhibited MSC migration and impaired the colonization of MSCs in neointima although extracellular HMGB1 as a chemotactic factor was highly expressed in allograft vessels. However, this is contrary to some previous studies that revealed that HMGB1 induced cell migration by forming a complex with CXCL12 and signaling via CXCR4, independent of RAGE pathway [38, 39, 40]. Indeed, these are not mutually exclusive; extracellular HMGB1 binds to RAGE to stimulate HMGB1 production, further activating RAGE and more HMGB1 production via a positive feedback loop  thus favoring HMGB1-CXCL12/CXCR4-mediated cell migration. Moreover, cell migration to HMGB1 requires nuclear factor κB (NF-κB) activation depending on the inhibitor of NF-κB kinase subunit beta. It is necessary for both CXCR4 and RAGE to maintain their sufficient expression , suggesting an intrinsic link between RAGE and CXCR4.
The ability of HMGB1 to stimulate MSCs was translated to a successful application in cell therapy for TA. MSC transplantation holds great promise for the treatment of arteriosclerosis by regulating the immune response as well as generating tissue repair [10, 11]. However, a series of recent studies has shown that vascular resident stem cells acquiring MSC-like phenotype migrated to the endothelium in response to vascular injury and differentiated to SMCs. The stem cells and their differentiated SMCs constituted a major cellular component of the hyperplastic neointima [13, 14, 42, 43, 44, 45, 46]. Although stem cells from the vascular walls and other sources might not definitively display a similar differentiation potential, it was suggested that transplantation of unselected or naive MSCs impairs the anti-atherogenic effect or even has an adverse effect. The present study demonstrates that HMGB1-modified MSCs possess a better anti-TA property. Neointimal thickness and macrophage infiltration were dramatically reduced along with a decline in vascular HMGB1 and RAGE expression after the recipients were transplanted with pLV-HMGB1-transfected MSCs. Possible mechanisms are as follows. Firstly, cell migration to allograft vessels was a prerequisite for vascular repair. HMGB1 promoted the migratory capacity of MSCs and stimulated cell recruitment to allograft vessels. The chemotaxis required activation of the HMGB1/RAGE axis since RAGE knockdown impaired MSC migration and produced little therapeutic effect. Secondly, HMGB1 preferably induced the differentiation of MSCs to ECs rather than to SMCs, although both PDGF and VEGF were highly expressed in allograft vessels (Additional file 5: Figure S4). On the one hand, it promoted endothelial regeneration following chronic rejection, which was evidenced by the Evans blue staining test showing that infusion of HMGB1-overexpressing MSCs improved graft endothelial integrity. Theoretically, ECs derived from infused MSCs were isogeneic to the host so as to induce immune tolerance by regenerating endothelium to replace original intima that was allogeneic and rejected by the host immunity. On the other hand, it prevented intimal hyperplasia by inhibiting the recruited MSCs to differentiate to SMCs. If the majority of MSCs differentiated to SMCs, neointimal formation would have been inevitably accelerated by an increased accumulation of SMCs and excessive deposits of extracellular matrix secreted by SMCs in the neointima. Despite its regulation on cell differentiation, HMGB1 slightly inhibited the viability of MSCs. This could threaten cell survival in vivo, but was presumably favorable for improving the outcome of MSC transplantation as the recruited cells were not allowed to proliferate excessively to accelerate intimal hyperplasia. Finally, the immunoregulatory ability of MSCs should not be neglected in cell therapy for TA. It was reported that MSC transplantation promoted the production of regulatory T cells and anti-inflammatory cytokines and inhibited activation of alloreactive immune cells [2, 3, 4, 5, 6, 7]. Our study revealed that infusion of HMGB1-overexpressing MSCs attenuated macrophage infiltration in graft neointima. Thus, it will be necessary to investigate the immunological changes after transplantation of HMGB1-modified MSCs in follow-up research.
The present study demonstrated that the HMGB1/RAGE axis is instrumental in regulating proliferation, migration, and endothelial differentiation of MSCs. HMGB1-modified MSCs were less proliferative than naive MSCs but acquired a better capacity for migration and endothelial differentiation, favoring their application in MSC transplantation for TA.
This study was made possible by grants from National Natural Science Foundation of China (81400306), Project of Medicine and Health Science and Technology of Zhejiang Province (2014KYA175), and Project of Invigorating Health Care through Science, Technology and Education of Jiangsu Province Medical Youth Talent (QNRC2016671).
National Natural Science Foundation of China (81400306). Project of Medicine and Health Science and Technology of Zhejiang Province (2014KYA175). Project of Invigorating Health Care through Science, Technology and Education of Jiangsu Province Medical Youth Talent (QNRC2016671). The authors take full responsibility for the study work, including design of the study and collection, analysis, and interpretation of data, and manuscript writing. The content is solely the responsibility of the authors and does not represent the official views of grant providers.
Availability of data and materials
The datasets generated and/or analyzed during the current study are available from the corresponding authors on reasonable request.
XM and MC carried out cell experiment studies and drafted the manuscript. WS and XT carried out the animal experiments. MS participated in the histological analysis. WW and BW participated in the design of the study, performed the statistical analysis, and gave a major revision to the manuscript. XZ conceived of the study and participated in its design and coordination. RY generously provided technical assistance in animal experiment. All authors read and approved the final manuscript.
All animals received humane care in compliance with the ‘Principles of Laboratory Animal Care’ formulated by the National Society for Medical Research and the ‘Guide for the Care and Use of Laboratory Animals’ prepared by the Institute of Laboratory Animal Resources and published by the National Institutes of Health (NIH Publication No. 86-23, revised 1996). The experiment was approved by the Committee of Animal Experiment Ethics at Nanjing Medical University.
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 9.Vanikar AV, Trivedi HL, Thakkar UG. Six years’ experience of tolerance induction in renal transplantation using stem cell therapy. Clin Immunol. 2017;95(1):155–60.Google Scholar
- 27.Vogel S, Borger V, Peters C, et al. Necrotic cell-derived high mobility group box 1 attracts antigen-presenting cells but inhibits hepatocyte growth factor-mediated tropism of mesenchymal stem cells for apoptotic cell death. Cell Death Differ. 2015;22(7):1219–30.CrossRefPubMedPubMedCentralGoogle Scholar
- 28.Vogel S, Chatterjee M, Metzger K, et al. Activated platelets interfere with recruitment of mesenchymal stem cells to apoptotic cardiac cells via high mobility group box 1/Toll-like receptor 4-mediated down-regulation of hepatocyte growth factor receptor MET. J Biol Chem. 2014;289(16):11068–82.CrossRefPubMedPubMedCentralGoogle Scholar
- 39.Di Maggio S, Milano G, De Marchis F, et al. Non-oxidizable HMGB1 induces cardiac fibroblasts migration via CXCR4 in a CXCL12-independent manner and worsens tissue remodeling after myocardial infarction. Biochim Biophys Acta. 2017;1863(11):2693–2704.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.